Abstract

When there is downtime in the cloud, two questions arise: is it a breach of a service-level agreement (SLA), and what must be done about it? We propose an architecture that connects telemetry with evidence assessment, parametric payment, and controlled recovery through a common incident record. Metrics are compared with a versioned policy, vigilant witnesses certify the evidence, and a smart contract enforces the principle of bounded payment. A separately authorized “director” can perform an allowed recovery operation if needed. The design deals directly with missing observations, duplicate claims, contradictions between witnesses, confidentiality, emergency suspension, and the danger of making a downtime worse through automated recovery. It also distinguishes the integrity of evidence from the fidelity of measurement. The paper is a design and evaluation plan; it reports no experimental results.

Keywords

How to Cite This Article

Rights